Effect of Air Jet Micronization on Particle Properties an the Correlation of Interparticle Interactions by Atomic Force Microscopy with Surface Forces by Inverse Gas Chromatography Vibha Puri, 1 Jageep Shur 2, Robert Price 2, Anreas Stumpf 1, & Ajit Narang 1 1 Small Molecule Pharmaceutical Sciences, Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080, USA 2 Nanopharm Lt., Cavenish House, Newport NP10 8FY, UK Summary Micronization of pharmaceutical powers can inuce structural an surface isorer on the surface of constituent particles, which can impact their aeroynamic performance in ry power inhaler (DPI) formulations. Furthermore, these materials unergo surface re-construction on storage, which can then impact their physicochemical properties over time. Ientifying an mapping the interparticle interactions that can be irectly linke to aeroynamic performance is challenging. In this work, we stuie the correlation of interparticle interactions measure by atomic force microscopy (AFM) with particle surface forces by inverse gas chromatography (IGC) for a moel active pharmaceutical ingreient (API), Compoun A. Micronize API prouce by air jet milling was store uner ifferent stress conitions. The rug-rug cohesive interactions an rug-lactose ahesive interactions were measure by AFM. The cohesive-ahesive balance (CAB) ratio was calculate. The freshly micronize API showe an AFM CAB ratio of 0.69 that suggeste greater APIlactose ahesive interactions, while its IGC CAB ratio of 1.16 suggeste higher API-API cohesive interactions. Nonetheless, both techniques reporte reucing CAB ratio upon storage of the micronize API at accelerate conitions of temperature an humiity, suggesting increase in API-lactose ahesive interactions an lowering of API cohesive interactions. The magnitue of change in AFM CAB ratios was greater than the IGC CAB ratios, inicating greater sensitivity of AFM technique. In summary, while the initial CAB may epen on the technique use for measurement, changes in the CAB on storage were consistent across the two analytical tools investigate. Introuction The performance of pharmaceutical powers in rug proucts, such as ry power inhalations (DPIs) an power mixtures for granulation, is preominantly governe by the interparticle interactions in the multi-component power blen 1. The surface properties of pharmaceutical powers are largely controlle by the processing history of the active pharmaceutical ingreient (API) 2. Air jet micronization processes are commonly use for particle size reuction of synthetic, crystalline APIs. However, this high energy process can inuce structural an surface isorer in the micronize API. Furthermore, this mechanically activate API on storage can unergo relaxation or surface stabilization that can alter the material s surface properties. Hence, there is a requirement for tools that expeite rigorous characterisation of the surface of micronize pharmaceutical solis. Surface-base techniques such as atomic force microscopy (AFM) an inverse gas chromatography (IGC) have been employe to map interparticle interactions 3, 4. These techniques are further use as preictive tools to unerstan prouct esign an performance. The sample preparation an the measurement principles iffer significantly for these two techniques 4. For example, the cohesive-ahesive balance (CAB) approach to colloiprobe AFM, employs use of crystallize substrate on which micronize particles are interacte by over a known area of the ominant crystal face of the API an excipient such as lactose 5. The force of cohesion an ahesion are then etermine from the array of force-istance curves. In contrast, the IGC technique measures surface free energy in bulk power sample by surface asorption of gases. Both techniques enable the investigation of the work of cohesion an ahesion of processe pharmaceutical solis. In this stuy, we characterize the cohesive an ahesive interactions of micronize Compoun A using AFM-CAB an IGC an investigate effect of storage of micronize API at high temperature an humiity conitions on the changes to the CAB. Materials Compoun A was micronize by air jet micronization (Foo Pharma Systems PM2 mill, Italy) with milling conitions of grin pressure of 7 bar an injection pressure of 8 bar. The micronize API was store at 25 C/60% RH an 40 C/75% RH open conitions for 8 weeks. Lactose monohyrate (Respitose SV003) was obtaine from DFE Pharma (NJ, USA).
Drug Delivery to the Lungs (DDL2017), 2017 - Effect of Air Jet Micronization on Particle Properties an the Correlation of Interparticle Interactions by Atomic Force Microscopy with Surface Forces by Inverse Gas Chromatography Methos Cohesive-ahesive balance (CAB) by colloi probe atomic force microcopy (AFM) Smooth lactose monohyrate crystalline substrates were prepare by a metho escribe by Begat et al. Unmicronise crystals of compoun A ha the require rugosity on which the force of cohesion was etermine. Five micronise rug particles at the initial time-point an upon storage uner ifferent environmental conitions were attache to AFM cantilevers using a custom-built micro-manipulation technique. AFM cantilevers with rug particles attache to them are referre to as rug probes. Cohesion an ahesion force measurements between the rug probes an primary crystals (rug or lactose monohyrate) were conucte using force volume moe (n= 5 probes), which recors a total of 1098 force-istance curves within a specifie area at 25 C an 44 % RH. The force of cohesion an ahesion for each rug probe for each API sample was plotte against each other to prouce a CAB plot. The regression analysis was conucte on Minitab, which provie an output of the stanar error of the regression an error of the estimate. The AFM-CAB measurements were then performe as escribe by Begat et al an Kubavat et al 6. Cohesive-ahesive balance (CAB) by inverse gas chromatography (IGC) IGC experiments were conucte on surface energy analyzer (SMS Instruments, Allentown, USA) with flame ionization etector. Power samples were packe into silanize glass columns (2 mm iameter) in weight range of 50-200 mg, tappe for 4 min, an visually inspecte for absence of voi spaces. Columns ware conitione for 2 h at 25 C an 0% RH conitions, followe by pulse injection measurements. Methane was use to etermine the column ea time. The probe solvents use were nonane, octane, heptane, hexane, ethyl acetate an ichloro methane (purity>99%, HPLC grae). Probe solvent were injecte to achieve target surface coverage (n/nm) in the range of 0.5 to 10% at 25 C an 0% RH conitions with helium carrier gas at flow rate of 10 ml/min. Surface coverage is efine as n/nm, where n is number of probe moles asorbe an nm is the number of moles require for a theoretical monolayer surface coverage. Samples were analyze in triplicate. The ispersive work of cohesion an work of ahesion plots were etermine using the ispersive surface energy of the API an lactose (mean of n=3). The IGC-base CAB ratios were calculate as the ratio of the average ispersive work of cohesion an ahesion values at infinite surface coverage of 0.04 n/nm an as slope of the plot of work of cohesion an ahesion for surface coverage range from 0.02 to 0.1 n/nm. IGC ata analysis The surface energy components were etermine from the retention time of ispersive an polar probes which interact with the soli surface sample. At each surface coverage, the net retention volume (VN) was calculate for each probe 7. When a series of liqui n-alkane are use as probes, the asorption ispersive free energy of the methylene group G CH 2 can be calculate from the slope of the plot of asorption free energy of the probes versus the carbon number n using the Dorris-Gray metho 8 G CH 2 = RT. ln ( V N,n+1 V N,n ) where R is the universal gas constant (J/mol K), T is the temperature (K), V N,n is the net retention volume of the n- alkane probe with the carbon number n. The ispersive surface energy of soli sample γ s can be obtaine by: γ s = 2 1 ( GCH2 ) 4γCH 2 N. a CH2 Surface energy is irectly relate to the thermoynamic work of ahesion between two materials. Accoring Fowkes 9 the total Work of cohesion/ahesion can be escribe as the sum of the ispersive contribution an the specific (or aci-base) contribution to the work. The ispersive work of cohesion (W Coh )an ahesion (W Ah ) can be calculate with the following equations W Coh = 2 (γ s ) W Ah = 2 (γ s. γ l ) 1/2
Results an Discussion The AFM base CAB plots (Fig 1) were obtaine using the force of cohesion an ahesion for the micronize API with lactose, for samples store uner ifferent conitions. The AFM CAB ratios are provie in Table 1. The AFM CAB ratio of micronize API batch was 0.69, which suggeste that the ahesive interactions of the API to lactose were 1.45 times greater than the cohesive interactions of the API particles. After 8-weeks of storage at 25 C/60% RH an 40 C/75% RH conitions the AFM CAB ratio reuce to 0.35 an 0.30, respectively. This suggeste about 2-fol increase in the API affinity to the lactose compare to the API cohesive interactions. Figure 1. AFM base force of cohesion an force of ahesion plots of freshly micronize API an after storage at ifferent conitions. The force of ahesion is etermine between API an lactose Table 1. Comparison of CAB ratios of ifferent samples erive from AFM an IGC techniques. Fig 2. shows the ispersive surface free energy plots for the freshly micronize API, an micronize API after storage uner ifferent environmental conitions. As the plots show, the SFE reuce significantly when store at 40 C/75% RH conition an to a smaller extent for samples store at 20 C/65% RH conition. The polar SFE showe less significant changes (ata not shown). Thus, changes in the surface interactions of the API coul primarily be relate to reuction in the ispersive surface energy. IGC-base CAB ratios inicate that the micronize API cohesive interactions lowere on exposure to the stress conitions (Table 1 an Fig 3). During storage uner high humiity conitions, water can evelop strong interactions (e.g., hyrogen boning) with the high energy sites (surface isorer/surface amorphous regions) of the micronize API an cause plasticization of these regions. This phenomenon of surface relaxation can then lower the surface energy of the particles. In case of micronize Compoun A, both, the ispersive an specific (polar) surface energy lowere, with overall increase in the contribution of specific surface energy to the total surface free energy. The AFM an IGC base CAB ratios for the freshly micronize API were significantly ifferent. The AFM CAB ratio of 0.69 suggeste greater rug-lactose ahesive interactions, while the IGC CAB ratio of 1.16 suggeste higher rug-rug cohesive interactions. As escribe by Bunker et al. 10, the AFM technique measures irect forces that are sum of the Van er Waals forces, electrostatic forces, an capillary forces uner the testing conitions. However, surface energy measurements by IGC are a measure of the Van er Waal forces only. AFM was performe at 44% RH conitions, which coul promote ahesive interactions between API an lactose, through the capillary forces 11. Aitionally, the freshly micronize Compoun A is likely to have electrostatic forces contributing to the surface interactions. In contrast, the IGC analysis was conucte in 0% RH environment, where there woul be absence of surface water, an the surface energy of API an lactose coul be ifferent when expose to 44% RH.
Drug Delivery to the Lungs (DDL2017), 2017 - Effect of Air Jet Micronization on Particle Properties an the Correlation of Interparticle Interactions by Atomic Force Microscopy with Surface Forces by Inverse Gas Chromatography Nonetheless, for the stresse API samples both techniques showe similar tren of reucing CAB ratio suggesting that the treate API particles ha greater affinity for API-lactose interactions, than the cohesive interactions. Although for the stresse samples, the extent of change in the AFM erive CAB ratios was higher than for the IGC erive CAB ratios. Interestingly, the IGC CAB ratios trens measure using the infinite regimen an a wier finite ilution regimen iffere, with the later showing more ifferences between the samples. Figure 2. Dispersive surface free energy plots of freshly micronize API, an micronize API store at 25 C/60% RH an 40 C/75% RH open conitions for 8 weeks (n=3 mean±s.d.). The compoun A was foun to be non hygroscopic in nature, with moisture uptake about 0.3% wt at 25 C-90% RH. Therefore, the change in the force balance was less likely ue to capillary forces but more likely ue to physical annealing of the material. From the formulation performance, the ata suggests that the micronize Compoun A is likely to exhibit less self-agglomeration an might achieve homogenous blen. Over time the ahesive interactions were observe to increase, which coul lea to changes in performance, for example reuce power ispersibility or aerosolization performance of DPIs. Figure 3. IGC base ispersive work of cohesion an work of ahesion plots of freshly micronize API, an micronize API store at 25 C/60% RH an 40 C/75% RH open conitions for 8 weeks. The work of ahesion is etermine between API an lactose over surface coverage of 0.02 to 1 n/n m.
Conclusions Ait jet micronization of the Compoun A significantly change its surface properties. Further, on storage, the micronize API unerwent surface relaxation that change its surface interactions. This stuy highlights that apart from the nee to fully unerstan the effect of micronization process variables, the storage history of the micronize API can also impact the overall prouct performance. These factors can be critical to the rug prouct performance, for example, by impacting the API istribution homogeneity at low rug loaing an the aeroynamic performance of ry power inhalation formulations. For the rug-lactose system stuie, the interparticle interactions etermine from AFM an IGC matche only in qualitative trens. The AFM an IGC CAB ratios were significantly ifferent for the freshly micronize API, however both techniques showe similar tren of increase API-lactose ahesivity uner stress conitions. As iscusse the two techniques iffer in the sample preparation, experimental conitions, an the kin of measurement mae an this coul contribute to the ifferent interpretations the two techniques provie. References 1. Xu Z, Mansour HM, Hickey AJ. Particle Interactions in Dry Power Inhaler Unit Processes: A Review, J Ahes Sci Technol 2011; 25 (4-5):pp451-482. 2. Rasenack N. Particle Engineering for Pulmonary Dosage Forms, American Pharmaceutical Review 2010; April 01, 2010. 3. Cline D, Dalby R. Preicting the Quality of Powers for Inhalation from Surface Energy an Area, Pharm Res 2002; 19 (9):pp1274-1277. 4. Jones M, Buckton G. Comparison of the cohesion-ahesion balance approach to colloial probe atomic force microscopy an the measurement of Hansen partial solubility parameters by inverse gas chromatography for the preiction of ry power inhalation performance, Int J Pharm 2016; 509:pp419-430. 5. Begat P, Morton DAV, Staniforth JN, Price R. The Cohesive-Ahesive Balances in Dry Power Inhaler Formulations I: Direct Quantification by Atomic Force Microscopy, Pharm Res 2004; 21 (9):pp1591-1597. 6. Kubavat H, Shur J, Ruecroft G, Hipkiss D, Price R. Investigation into the influence of primary crystallization conitions on the mechanical properties an seconary processing behaviour of fluticasone propionate for carrier base ry power inhaler formulations, Pharm Res 2012; 29 (4):pp994-1006. 7. Thielmann F, Burnett D, Heng J. Determination of the surface energy istributions of ifferent processe lactose, Drug Dev In Pharm 2007; 33 (11):pp1240-1253. 8. Dorris GM, Gray DG. Asorption of n-alkanes at zero surface coverage on cellulose paper an woo fibers., J Colloi Interf Sci 1980; 77 (2):pp353-362. 9. Fowkes F. Attractive forces at interfaces., In Eng Chem Chem 1964; 56:pp40-52. 10. Bunker M, Davies M, Roberts C. Towars screening of inhalation formulations: measuring interactions with atomic force microscopy, Expert Opin Drug Del 2005; 2:pp613-624. 11. Price R, Young PM, Ege S, Staniforth JN. The influence of relative humiity on particulate interactions in carrier-base ry power inhaler formulations, Int J Pharm 2002; 246:pp47-59.